The Bradley Effect, Redux?

How helpful are the reassessments of the much-debated theory? Not very.

The New York Times and The Washington Post both printed articles this weekend on the Bradley Effect and its forecasted impact on this election.

The Bradley Effect, of course, is the theory that black politicians are likely to do significantly worse in elections than exit polls would indicate (by as much as 6 or 7 percent), and is named after Tom Bradley, the former mayor of Los Angeles who narrowly lost a gubernatorial race after leading in the polls. (It’s otherwise known as the Wilder Effect or Dinkins Effect, after L. Douglas Wilder and David Dinkins, two black politicians for whom this discrepancy seemed to hold.) The existence (and relevance today) of this much-studied phenomenon is debated enough that, in a column last month, William Safire suggested putting a categorical “‘so-called’ in front if you dispute it.”

It’s no surprise that both papers chose to run these articles—the NYT piece, by Kate Zernike, in its Week in Review section, and the Post piece, by Stephen Holmes, appearing on page A06—given the growing question mark of exactly how large a factor race (and racial prejudice, and subconscious racism, and “racism without racists” etc.) will be in this election. This reassessment of the Bradley Effect (which is really just a re-focusing of attention on it) clearly comes on the heels of a week characterized by sharp and divisive rhetoric on the campaign trail.

What’s most interesting about the two articles is how formulaic they are: they cite many of the same sources—a slew of pollsters and political scientists—and both concede early on that there is no solid answer to the question that they pose. (The Times’s take is that “even pollsters say they can’t be sure how accurately polls capture people’s feelings about race,” while the Post boasts the rather skeptical title: “Pollsters Debate ‘Bradley Effect’”.)

Of the two writers, it is Holmes who more clearly outlines the current debate—whether the Bradley Effect is an outdated theory—by offering twin reasons for its possible obsolescence: “…if the effect has disappeared, it is not clear whether that is because polling techniques have improved or because the country has become more tolerant about race.”

But the only way to be clear, it seems, is to frankly state that everything—what matters, what doesn’t, what may seem to matter but will actually turn out not to—is unclear. The rest of Holmes’s article follows this pattern, in what amounts to a preponderance of uncertainty: “Most experts say they do not believe that the phenomenon, known as the “Bradley effect,” is at work in this election. But some disagree.”

Also: “But whether [California gubernatorial candidate Tom] Bradley lost because of hidden racism has never been clear” and “Finding hard evidence for or against a Bradley effect today is difficult.”

And again: “Experts agree that it is often difficult to fully tease out the extent to which race plays a factor in voting decisions.”

The experts cited vary widely in their opinions, too, and their responses form an open-ended composite. Michael Dawson, a professor of political science at the University of Chicago, “remains skeptical about the willingness of whites to vote for a black candidate – and the ability of polling to capture that reluctance.” Anthony Greenwald, a professor of psychology at the University of Washington, “does not buy that people are lying to pollsters,” but says polls are inaccurate because undecided voters make up impromptu responses. Andrew Kohut, the president of the Pew Research Center, “theorizes that polling discrepancies do not come from respondents who lie, but from people who decline to participate in polls.”

Holmes ends on a sincere, but weak, note: “…this presidential election will be a highly visible test of just how real [the Bradley Effect] is.” We don’t know yet, but this election will be a good indicator of how relevant this theory is today.

Zernike’s article is similar, arguing that the Bradley Effect “obscures…the more important point: there are plenty of ways that race complicates polling.” (These include the “reverse” Bradley Effect, wherein black candidates perform better than exit polls would indicate, and the existence of “I don’t know” voters, wherein those who fear being perceived as racist say instead that they don’t know who they will vote for.)

So we don’t know. We can’t. Experts disagree. And while that’s a legitimate strain of inquiry, it seems a bit predictable. These articles feel, inevitably, like an indirect response to the volatile rhetoric that we witnessed all week—i.e. a responsible, journalistic way to respond to the unchecked vitriol.

Without a doubt, the hate speech and signage directed towards Obama during recent Republican rallies are worrisome. And they’re dangerous for their consequences—as Frank Rich writes, “unleash[ing] the demons who have stalked America from Lincoln to King.” The Post’s Jonathan Capehart wrote last Friday about “an anger and resentment on the campaign trail that should leave all with a cold chill running through their bodies.” A homemade sign at a McCain rally today read, incredibly, “Obama bin Lyin’”. This is a topic that should be—and is being—addressed.

But it’s also simplistic, and problematic, to advance an evergreen theory like the Bradley Effect at times like this, because it suggests (even unintentionally) that the best way to talk about race-in-politics is still to cite surveys and debate the efficacy of exit polls—in other words, to talk about it in the understood and studied ways. While turning theories around and around may be a safe and necessary way to address the extent and effect of racism in this campaign, their prescriptive “despite experts, we still can’t say for sure” formulations don’t actually do anything for anyone’s cold chills.

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Jane Kim is a writer in New York.